Malaysian experts: Why iPhone X’s Face ID may have just solved the ‘security friction’ problem
Face ID – During an interview with Computerword, CF.Fong, Founder of LGMS said that:
“Using Biometric lingo, facial recognition is less ‘Intrusive’ and less ‘disruptive’ when compared to other biometric technologies available now, such as voice recognition, fingerprint scanning, retina scanning, etc.,” he said.
“Essentially, the effectiveness of biometric devices is measured by the likelihood of having false positives (errors), It’s rather challenging for facial recognition technologies to acquire a consistent high accuracy rate, especially if you compare it with other biometric devices like fingerprint or retinal scanners.”
“Also, because of ageing and life styles, our facial appearance may change over time whereas our fingerprints and retina patterns will not have any change due to ageing,” Chong said.
“The good news is that with the emergence and integration of today’s advanced technologies -such as AI, artificial neural networks, and machine learning – facial recognition technology has now become more reliable and deliver consistent accuracy,” he said. “For example, machine learning can help computers to ‘predict’ and ‘anticipate’ what we will look like when we age, or when we are wearing a spectacles or make-up.”
“With the convergence of technologies, we can expect to witness the boom of biometric technologies in next the two years,” Fong said.
Dr Amirudin agrees. “Face identify is now a universal and common data set, which allows rapid identification and verification. For facial recognition systems to work accurately, several poses are usually required to ensure all angles of a face is ‘trained’ by the system’s machine learning algorithms (AI). In addition mobile biometrics applications, face ID is useful for public safety. The methodology is very useful in forensics and automatic detection of individuals blacklisted by authorities.”
Article from ComputerworldMalaysia.
You may also find other interesting LGMS News at the following links: